Research on class size effects: a critique of methods and a way forward

Research findings on class size differences need to be examined carefully because of their far reaching pedagogical and resource implications. An examination of research on the possible causal effect of class size differences on pupil progress, and on possible mediating factors, also raises important issues for educational research in general. In this chapter critically different approaches to research on class size effects are reviewed and an approach used in current research at the Institute of Education, that seems most likely to inform this controversial topic is described. Researchers need to pay more attention to providing a more reliable and valid measure of class size itself. In keeping with the bulk of research in this area the emphasis is on quantitative approaches, although possible limitations and alternative research approaches are considered. Limitations of earlier cross sectional research are examined, and the well-known STAR research is used to highlight some limitations of experimental designs. Finally, the advantages of longitudinal research, using multi-level modelling techniques, are examined.